The LC Handbook - Conquer Scientific · Handbook Guide to LC Columns and Method Development. The LC...

of 163/163
The LC Handbook Guide to LC Columns and Method Development
  • date post

  • Category


  • view

  • download


Embed Size (px)

Transcript of The LC Handbook - Conquer Scientific · Handbook Guide to LC Columns and Method Development. The LC...

  • The LC HandbookGuide to LC Columns and Method Development

  • The LC HandbookGuide to LC Columns and Method Development

  • 2

    Introduction 4

    Essential chromatography concepts 5Efficiency (N) 6

    Retention Factor (k) 7

    Selectivity or separation factor (α) 7Resolution (Rs) 8

    Pressure 9

    van Deemter Curves 10

    The gradient equation 10

    Selecting your HPLC column 12HPLC mode 13

    Column selection basics: conventional columns 14

    • HighPerformanceLiquidChromatography(HPLC)columns 15

    • UHPLCcolumns 16

    • Superficiallyporousparticlecolumns 16

    • ColumnsforLC/MS 17

    • ColumnsforGelPermeationChromatography(GPC),SizeExclusionChromatography(SEC),andGelFiltrationChromatography(GFC) 18

    • Columnsforbiocharacterization 18

    Column characteristics 19

    • Silica 19

    • Bondedphases 19

    • Polymers 19

    • Poresize 20

    • Particlesize 20

    • Columndimensions 21

    Cartridge column systems 22

    Keys to performance: column configurations and settings 25The importance of reducing extra-column volume 26

    Preparing the perfect fitting connection 27

    • Makingagoodconnection 28

    Sample injections 30

    Setting the data collection rate 31

    Dwell volume and its impact on chromatography 32

    • Measuringyoursystem’sdwellvolume 34

    • Evaluatingtheimpactofdwellvolume 36

    • Dwellvolumeandanalysistime 37

    Chelating compounds 37

    pH and mobile phase modifiers 38

    Working with gradients 39

    Optimizing column re-equilibration 40

    Column aging 42

    • Lossofbondedphase 42

    Cleaning a reversed-phase silica column 42

    Cleaning a normal phase silica column 44

    Cleaning a reversed-phase polymeric column 44

    Method development 45Method development: where to start 46

    • ModeSelection 46

    • Choosingthecolumnandpackingdimensions 49

    • Choosingthestationaryphase 50

    Method development for reversed-phase chromatography 50

    Selection of stationary phase for reversed-phase chromatography 50

    Selection of mobile phase solvents for reversed-phase chromatography 53

    • Workingwithmobilephases 53

    • Troubleshootingmobilephasesandmobilephasemodifiers 53


  • 3

    • Mixingmobilephases 54

    • Degassingmobilephases 54

    Managing your pH with mobile phase modifiers 54

    • CommonbuffersforUVdetectors 56

    • ConsiderationsforLC/MS 57

    • Troubleshootingissueswithmobilephasemodifiers 59

    • Troubleshootingexample:driftingbaseline 59

    • Troubleshootingexample:broadeningorsplittingcausedbyhighpH 61

    • Troubleshootingexample:mobilephasemodifiersandselectivity 61

    Optimizing your chromatographic conditions for reversed-phase chromatography 63

    • Isocraticoptimization 64

    • Gradientoptimization 67

    Polymeric columns for reversed-phase chromatography 69

    A step-by-step guide for 'hands-on' isocratic method development in reversed-phase chromatography 70

    Tips for transferring methods from conventional columns to high efficiency columns 72

    Automated method development tools 74

    Method development for other HPLC modes 76

    • HILIC 76

    • Normalphasechromatography 78

    • Ion-exchangechromatography 80

    • Gelpermeationchromatography/sizeexclusionchromatography 81

    Protecting your chromatographic results 82Sample Preparation 83

    • Samplefiltration 83

    • SolidPhaseExtraction 84

    • Liquid-LiquidExtraction(LLE) 85

    • SupportedLiquidExtraction(SLE) 86

    • QuEChERS 87

    The importance of using high-quality-grade solvents 88

    Special considerations for UHPLC 88

    Inline filters 89

    • Low-volumeinlinefilters 89

    Guard columns 90

    Solvent-saturation columns 90

    Column inlet frits 90

    Column care and storage 90

    • Maximizingcolumnlifetime 90

    • Careinstorage 92

    • Unblockingacolumn 92

    Quickly determining when a column is going bad 92

    Ensuring method reproducibility around the world 93

    • Increasingmethodrobustness 93

    • Areminderaboutdwellvolumeimplications 94

    Quick troubleshooting reference 95Tips for effective troubleshooting 95

    Useful references 101USP designations 101

    Solvent miscibility 107

    UV cutoffs for mobile phase modifiers 110

    Solid Phase Extraction sorbents 111

    SPE sorbent conditions 112

    Other suggested reading 113

    Other Agilent resources 113

    Glossary 114

    Index 141

    Agilent Products and ordering information 147

  • 4


    Where to begin? Liquid chromatography is a vast and complex subject, but one for which we never lose our interest.

    Chromatographers around the world are using HPLC techniques to ensure the safety of our food and water, develop life-saving pharmaceutical products, protect our environment, guard public health, and that’s just the start of it. The more you know about chromatography, the more you can get done with this amazing technology.

    Today, you have more choices of columns and packing materials to suit an ever expanding range of uses. Agilent now offers more than 2,000 column choices covering the broadest array of applications and conditions. This increases your opportunities to select the most appropriate column for your needs.

    As part of our commitment to helping you get the best results from your liquid chromatography, we have compiled this handy guide to choosing LC columns, with plenty of tips and tricks to make your job easier and more productive. In addition, we’ve drawn on more than 40 years of experience to provide suggestions for overcoming some of the common problems that can occur with columns and fittings in everyday use. The guide covers the main columns used in LC, with particular emphasis on reversed-phase high performance liquid chromatography.

    How to use this guide:

    • Sections are color-coded for your easy reference.

    • The glossary in the back is fairly comprehensive. It’s intended to be a good resource, although we have not touched on every glossary term in the rest of the book, due to space considerations.

    • This book primarily focuses on reversed-phase HPLC although we highlight other techniques elsewhere in the book.

  • 5

    Essential chromatography concepts

    We all remember the feeling we had in school as we learned math, wondering how it would actually come into practical use. Scientists have to learn more math than many professionals, and this section reminds us why.

    Here, we will briefly review the equations and theory behind many of the concepts that drive chromatography. Understanding these concepts will help you to get the best results, and to troubleshoot if you encounter problems.

    We start with fundamentals of performance:

    • Efficiency

    • Retention

    • Selectivity

    • Resolution

    • Pressure

    These are all key to understanding how to optimize results and successfully develop methods.

    We also explore a few more complex concepts:

    • van Deemter curves

    • The gradient equation

    These two topics are also important for method development.

  • 6

    Columns with high plate numbers are more efficient. A column with a high N will have a narrower peak at a given retention time than a column with a lower N number.

    Efficiency (N)

    Column efficiency is used to compare the performance of different columns. It is probably the most frequently cited parameter of column performance and is expressed as the theoretical plate number, N.

    EfficiencyN = 16 (tR/wt)

    2 For Peak B, 16(4.5 min./0.9 min.)2 = 400 platesk = (tR-t0)/t0

    RetentionkA = (2.5 - 1)/1 = 1.5kB = (4.6 - 1)/1 = 3.6kC = (6.2 - 1)/1 = 5.2

    Selectivity (C-B)α = k2/k1α = kc/kb = 5.2/3.6 = 1.44α = 1.44

    Selectivity (B-A)α = k2/k1α = kb/ka = 3.6/1.5 = 2.4α = 2.4

    Equation 1. Efficiencyequation

    Equation 2. Alternateequationforcalculatingefficiency

    Figure 1. Chromatographicillustrationofefficiency,RetentionFactorandresolution

  • 7

    If we measure the distance tw here (Figure 1), by drawing tangent lines to approximate the four-sigma peak width, we can measure the theoretical plates for peak B, using Equation 1, N = 16 (tR/tW)

    2. Sometimes the four-sigma peak width is difficult to measure (e.g., with a noisy baseline), so an alternate equation (Equation 2) involves measuring the peak width at half-height (w1/2): N= 5.54 (tR/w1/2)


    High column efficiency is beneficial since less selectivity is required to completely resolve narrow peaks. Column efficiency is affected by column parameters (diameter, length, particle size), the type of eluent (especially its viscosity), and flow rate or average linear velocity. Efficiency is also affected by the compound and its retention. When comparing columns, the number of theoretical plates per meter (N/m) is often used. However, the same chromatographic temperature conditions and peak retention (k) are required for the comparison to be valid. On stationary phases where α is small, more efficient columns are beneficial.

    Retention Factor (k)

    Formerly referred to as capacity factor or k´ (k prime), the Retention Factor measures the period of time that the sample component resides in a stationary phase relative to the time it resides in the mobile phase. It is calculated from the Retention Time divided by the time for an unretained peak (t0).

    Selectivity or separation factor (α)

    The separation factor is a measure of the time or distance between the maxima of two peaks. If α = 1, the two peaks have the same retention time and co-elute.

    Equation 3. RetentionFactorequation

    Equation 4. Selectivityequation

  • 8

    Selectivity is defined as the ratio in capacity factors. In Figure 1, you will see that there is better selectivity between peaks A and B than between B and C. Calculations are provided to demonstrate.

    Selectivity can be changed by changing the mobile phase constituents or changing the stationary phase. Temperature may also be a factor in adjusting selectivity.

    Resolution (Rs)

    Resolution describes the ability of a column to separate the peaks of interest, and so the higher the resolution, the easier it is to achieve baseline separation between two peaks. Resolution takes into consideration efficiency, selectivity and retention, as can be seen in Equation 5. One can improve resolution by improving any one of these parameters.

    In Figure 2, we see the different effects of each component on the separation process. All of these terms show a diminishing return. This means that the more you try to work on something to improve the separation, the less effective it will become.

    If you double the column length, you will obtain more theoretical plates, but your separation will take twice as long; you will only get a square root of 2, or 1.4 improvement in the resolution.

    A value of 1 is the minimum for a measurable separation to occur and to allow adequate quantitation. A value of 0.6 is required to discern a valley between two equal-height peaks. Values of 1.7 or greater generally are desirable for rugged methods. A value of 1.6 is considered to be a baseline separation and ensures the most accurate quantitative result.

    Equation 5. Resolutionequation

  • 9


    The Pressure equation (Equation 6) identifies five key factors that affect system pressure: solvent viscosity(h), flow rate (F), column length (L), column radius (r) and particle diameter (dp). It is a good idea to familiarize yourself with the pressure equation to understand these key contributors to system pressure.

    As noted in the formula, even a small decrease in the particle size (dp) has a significant impact on backpressure.

    Equation 6. Pressureequation

    Figure 2. Resolutionasafunctionofselectivity,columnefficiencyorretention

  • 10

    We often plot van Deemter curves to evaluate the performance of different columns, and to understand the optimum linear velocity (uopt) for a method.

    The gradient equation

    Whenever your sample has a wide variety of components present, it can be difficult to separate all of the components in a reasonable time using isocratic elution (e.g. constant mobile phase composition). Gradient elution is a process to increase the mobile phase strength as a function of time, resulting in faster analyses and better peak shape and quantitation. With gradient elution, peak widths are typically more narrow and of constant width (see p. 39 for more about gradients).

    The gradient equation (Equation 8) shows key variables that affect your analysis, and may cause issues with your chromatography if you don’t account for them. The equation shows how the retention factor is influenced by flow rate (F), gradient time (tG), gradient range (DF), and column volume (VM). It’s important to

    Figure 3. IllustrationofthevanDeemterequation

    Equation 7. vanDeemterequation

    van Deemter Curves

    The van Deemter equation evaluates efficiency (expressed as H, see Equation 7) as a function of linear velocity (u) or flow rate. The H – called plate height, or height of a theoretical plate – is determined by dividing the column length (L) by the calculated number of theoretical plates. The goal is to get a small plate height. We can do this most effectively with smaller particle columns, optimum linear velocities and low viscosity mobile phase. As particle size decreases, the optimum linear velocity increases.

  • 11

    In the gradient equation, S is a constant and is dependent on the size of the molecule being separated. For small molecules, the value of S is about 4 to 6. For peptides and proteins, S lies between 10 and 1,000.

    These days, it is common to change the dimension of the column, either to something shorter (e.g. for higher throughput) or with a narrower internal diameter (e.g. for mass spectrometric detection). Any decrease in column volume must be offset by a proportional decrease in gradient time (tG) or flow rate (F). Any change in the gradient compositional range (DF), using the same column, needs to be adjusted by a proportional change in gradient time (tG) or flow rate (F) if you want to maintain the same gradient slope and k* value.

    Equation 8. Gradientequation

    A great way to get help when transferring a method to a column with different dimensions is to use the AgilentMethodTranslationSoftware. You can find the method translator by typing in 'LC Method Translator' in the search field at

    remember that in order to keep the retention factor constant, changes in the denominator need to be offset by proportional changes in the numerator, and vice versa.

    Increasing the retention factor k (or k*, in a gradient) is an easy way to increase resolution, but as shown in Figure 2, it is not as effective as increasing efficiency or selectivity. If the retention factor is increased by increasing gradient time, you will have a longer run time, as Equation 8 shows.

  • 12

    Selecting your HPLC column

    Selecting a column is part science and part art. We say this because there are certainly some selection criteria that are straightforward: column phase for application, appropriate sizes for your system pressure limits, etc.

    In many cases, a specific phase or column dimension may be specified for a pre-defined method, so your selection will mostly have to do with meeting these specifications and ensuring a reliable, high quality column provider (we recommend Agilent, but we admit a bias).

    Beyond that, your method development may enable you to get more artful, in that you will want to test the performance of various stationary phases and mobile phases to get just the selectivity your application demands.

    This section will first review a few basics:

    • HPLC modes

    • Common HPLC column materials

    Then we'll talk specifically about the key types of columns most often used today and what makes them each unique.

    • HPLC columns

    • UHPLC columns

    • Superficially porous particle columns

    • Columns for LC/MS

    • Columns for Gel Permeation Chromatography (GPC), Size Exclusion Chromatography (SEC) and Gel Filtration Chromatography (GFC)

    • Columns for biochromatography

  • 13

    See pp 46 - 48 for more information about the differences between the various modes.

    We'll also explain the importance of key column characteristics

    • Packing – silicas, bonded phases, polymers

    • Pore size

    • Particle size

    • Column dimensions

    At the end of this section, we will provide an overview of cartridge-style columns that are available, including guard, semi-preparative, prep and process columns.

    The information in this section should be used in conjunction with the column selection information in the Method Development section (p. 45), which begins to incorporate more information about column stationary phases and selectivities.

    HPLC mode

    The HPLC mode is the most important factor to decide before you can begin evaluating columns. There are a number of common HPLC modes or techniques in use today:

    • Reversed-phase

    • HILIC

    • Normal phase

    • Ion-exchange

    • Size exclusion

    Generally, surveys have shown that 95% of all chromatographers use reversed-phase chromatography sometime during their work, and, for this reason, this booklet will focus mostly on this technique. We’ll touch on some of the others, however, to point out key differences and propose good resources for more information.

  • 14

    Column construction

    Column type Column internal diameter (mm)

    Particle type

    Particle size (µm)



    Stainless steel Analytical 4.0-4.6 Silica, polymer

    1.8-10 Traditional quantitative analysis

    Analytical solvent saver

    3 Silica 1.8-5 Reduced solvent consumption

    Analytical narrow bore

    2.0-2.1 Silica 1.8-5 Reduced solvent consumption

    Analytical microbore

    1 Silica, polymer

    3-5 Increased sensitivity, sample size ng to µg

    Analytical capillary

    0.3-0.5 Silica, polymer

    3-5 Sample size pg to ng

    Analytical nano 0.075-0.1 Silica, polymer

    3 Sample size < 1 pg

    Prep or semi-prep 9.4-50 Silica, polymer

    5-50 Sample purification

    Column selection basics: conventional columns

    Columns for liquid chromatography are made from cylinders of stainless steel or polymers or, more rarely, glass, containing bonded silica or polymer particles. They are available in many dimensions to suit the different needs of chromatographers and their applications. They range from short, narrow-bore columns for high throughput LC/MS, to 50 mm internal diameter (id) preparative columns for gram-scale purification, all the way up to preparative column packing stations with dimensions up to 600 mm for pilot, scale-up and production facilities. Column dimensions affect sensitivity and efficiency, and determine the amount of analyte that can be loaded onto the column. For example, small id columns improve sensitivity compared to larger id columns, but with reduced loading capacity.

    Modern stainless steel analytical columns (1 to 4.6 mm id, 20 to 250 mm long) use very low, or zero, dead-volume fittings. The column packing material is held in place by stainless steel frits at both ends.

    Figure 4. Diagramofacolumn

    Table 1. Somecharacteristicsofcolumnsforliquidchromatography Continuedonnextpage

  • 15

    Column construction

    Column type Column internal diameter (mm)

    Particle type

    Particle size (µm)



    Stainless steel Analytical 3 Silica 1.8-5 Good choice to match flow rate capability of Agilent MS

    Analytical narrow bore and microbore

    1-2.1 Silica, polymer

    1.8-10 Where higher sensitivity or specialty detectors are required

    Fused silica Analytical capillary

    0.3-0.5 Silica, polymer

    1.8-10 With MS detectors


    Stainless steel Analytical 4.6 Silica 1.8-2.7 For all fast

    and high



    Analytical solvent saver

    3 Silica 1.8-2.7

    Analytical narrow bore

    2.1 Silica 1.8-2.7


    Stainless Steel prep 10-100 polymer 10-50 Compound isolation


    Stainless Steel process 10 cm to 100 cm polymer 10-50 Compound production

    Note: Certain bioHPLC columns are available in PEEK for a metal-free sample path.

    Use ZORBAX Rapid Resolution High Throughput (RRHT) 1.8 µm columns and Poroshell 120 2.7 µm columns , up to 600 bar. Use ZORBAX Rapid Resolution High Definition (RRHD) columns, 1.8 µm, up to 1200 bar.

    High Performance Liquid Chromatography (HPLC) columnsSilica gel is commonly used as a stationary phase in normal phase, adsorption HPLC, and is the support for numerous chemically bonded stationary phases. The surface of the silica is covered with strongly polar silanol groups that interact with molecules in a non-polar mobile phase, or serve as reaction sites for chemical bonding. Normal phase HPLC works well with analytes that are insoluble in water, and organic normal phase solvents are more MS ‘friendly’ than some of the typical buffers used in reversed-phase HPLC. However, the technique sometimes suffers from poor reproducibility of retention times because water or protic organic solvents (which have a hydrogen atom bound to an oxygen or nitrogen atom) change the hydration state of the silica. This is not an issue for reversed-phase HPLC, which has become the main HPLC technique. In reversed-phase chromatographic systems, the silica particles are chemically modified to be non-polar or hydrophobic, and the mobile phase is a polar liquid.

  • 16

    Superficially porous particle columnsSuperficially porous particle (SPP) columns have enjoyed a recent resurgence in smaller particle sizes than the older 'pellicular' particle columns. As depicted in Figure 6, the Agilent Poroshell 120 particle has a solid core (1.7 µm in diameter) and a porous silica layer (0.5 µm thickness) surrounding it. The current interest in this technology is driven by its re-introduction in smaller particle sizes, such as the sub-3 µm sizes, for use in typical small molecule reversed-phase separations.

    The Poroshell 120 offers significant method development advantages to chromatographers using conventional totally porous columns. Because diffusion only occurs in the porous outer shell, not the solid core, efficiency is increased compared to a totally porous particle of the same size. In fact, a 2.7 µm SPP will give efficiency comparable to a 1.8 µm totally porous particle. A big advantage is the fact that the backpressure created by the SPP column is greatly reduced due to its larger particle size, allowing chromatographers to increase flow rate and improve the speed of their analysis, while enjoying exceptional resolution. It is also important to know that Poroshell 120 columns are packed with a standard 2 µm frit, so they are more forgiving for dirty samples and do not clog as readily as columns with smaller frits. Sample filtration, especially when using 1.8 µm columns, will also aid in reducing clogging.

    UHPLC columnsUltra High Pressure Liquid Chromatography (UHPLC) generally refers to liquid chromatography performed at pressures in excess of 400 bar (6000 psi), which was the conventional maximum system operating pressure for decades. Generally, UHPLC columns contain small particles (

  • 17

    The separation of peptides and proteins is challenging because they diffuse slowly, so flow rates must be kept low to prevent peak broadening. Agilent’s Poroshell 300 columns use a superficially porous 5 µm particle made with a thin layer of porous silica (0.25 µm thickness) surrounding an impervious solid-silica core. This technology reduces the diffusion distance, permitting rapid HPLC separation of peptides and proteins from 500 Da to 1,000 kDa.

    Columns for LC/MSThere are many columns for LC/MS, depending on the sample. For simple analytical samples, use short columns (with high resolution) to reduce analysis time for high throughput LC/MS. For higher resolution, use longer columns.

    Flow rate also affects your choice of a column. LC/MS systems typically operate at flow rates from 1 µL/min to 1 mL/min. This makes smaller id columns such as Agilent Solvent Saver (3.0 mm id), narrow bore (2.1 mm id), and capillary and nanobore columns (see Table 1, p. 15) good options for high sensitivity and fast analyses.

    The best bonded phase choice is a high performance end capped C18 bonded phase, stable over a wide pH range, compatible with the typical volatile mobile phase additives used for LC/MS, including formic acid and acetic acid.

    Figure 6. PoroshellParticles

  • 18

    Columns for Gel Permeation Chromatography (GPC), Size Exclusion Chromatography (SEC), and Gel Filtration Chromatography (GFC)Separations based on molecular size in solution are generally performed by size exclusion chromatography (SEC). This chromatographic mode is used to separate polymers, including biopolymers. It is used to characterize the molecular weight and molecular weight distribution of polymers. The columns are usually stainless steel and contain gels, cross-linked polymers, or silica particles with tightly controlled pore sizes. The separation mechanism relies solely on the size of molecules in solution, and unlike the other chromatographic modes, there should be no chemical interactions between the particles and the stationary phase. We use the term gel permeation chromatography (GPC) to describe the analysis of organic polymers (e.g. plastics) in organic solvents, and aqueous SEC to describe the analysis of water-soluble organic polymers (e.g. polyvinylalcohol) and water-soluble biopolymers (e.g. proteins, nucleic acids and polysaccharides) using predominantly aqueous mobile phases. Gel filtration chromatography (GFC) has been historically used to describe the low pressure separation of biopolymers such as proteins.

    For a comprehensive guide to GPC/SEC we recommended Agilent’sIntroductiontoGelPermeationChromatographyandSizeExclusionChromatography. (Agilent publication number 5990-6969EN)

    Columns for biocharacterizationBiochromatography columns, or biocolumns, are columns for the separation of biological compounds such as proteins and peptides, oligonucleotides and polynucleotides, and other biomolecules and complexes, including virus particles. Biocolumns are designed to greatly minimize or eliminate irreversible or non-specific binding of the sample to the packing and to retain biological function (enzymatic activity). Frequently, biocolumns are constructed so that active metals do not contact the sample. They may be made with polymers (e.g. PEEK), fused silica and glass-lined stainless steel, or metallic components that are coated to render the column biocompatible.

  • 19

    Column characteristics

    SilicaSilica is an ideal material for chromatography, and has been the main base packing material for bonded phase HPLC columns for decades. Silica particles are rigid and resist compaction due to flow, particularly important when particle sizes are 5 µm or less and higher pressures are necessary. Their extremely large surface areas provide the adsorptive capacity for HPLC and UHPLC, and the silanols or Si-OH groups on the particle surface are ideal bonding sites for functional carbon chains.

    Not surprisingly, spherical silica particles used for column packing are available in a variety of sizes, purity and acidity. Most modern packings are Type B silicas, which are very low in trace metals and are less acidic than older Type A silicas. Less acidic silica means less potential for interaction between a basic analyte and silanol groups on the silica surface, which contributes to improved peak shape. Columns made with high purity silica are the most common choices for today’s chromatographer.

    Bonded phasesFor reversed-phase chromatography, the most common bonded phase is a C18 (octyldecylsilane, ODS). This is just one of the many types of alkyl or carbon chains attached to the surface of the silica particle. Some other popular choices for linear alkylsilane phases include C8 and C4. Phenyl, including Phenyl-Hexyl and Diphenyl, and AQ, CN, and PFP phases can offer significant differences in selectivity from the straight-chain alkyl phases and may provide a successful separation. There are an increasing number of bonded phase choices with some specific to high aqueous mobile phases or other applications.

    In general, larger solutes, such as proteins, are best separated on short-chain reversed-phase columns (C3, CN, diol) bonded to wide pore silica gels (pore size: 300Å). Peptides and small molecules are separated on longer-chain columns (C8, C18). There are many cases, however, where this convention does not apply. Therefore, it is a good idea initially to select a phase in the middle of the hydrophobic spectrum (for example, C8), then change to a more hydrophobic phase or more hydrophilic phase depending on initial results and the solubility of your sample.

    There are other options to consider for ion-exchange, gel permeation/size exclusion or HILIC chromatography. We discuss these further on pp. 76 - 81.

    PolymersWhen a column is needed that can operate at very low and high pH, polymeric packings provide an alternative to silica-based materials. Polymeric particles are ideal for small-scale chromatography, particularly LC/MS, as they are chemically stable and do not leach soluble or particulate species. Reversed-phase spherical polymeric packings used in PLRP-S columns, for example, are based on a styrene/divinylbenzene copolymer with an inherently hydrophobic surface. No bonded phase is required for reversed-phase chromatography with polymeric particles. These rigid macroporous particles can be coated and derivatized to give a range of functionalities, including weak and strong cation and anion-exchangers.

  • 20

    Pore sizeThe choice of the pore size is determined by the molecular weight of the component which is analyzed. For reversed-phase separations of small molecules, choose a column packing with small pores (60-120Å). For small molecules and peptides use 100-150Å, for polypeptides and many proteins choose 200-300Å, and 1,000Å and 4,000Å for very high molecular weight proteins and vaccines.

    For GPC/SEC separations the molecular weight range for separations is typically given with the pore size information so the right column can be selected. Tables of this information are listed with the column choices.

    Particle sizeThe standard particle size for HPLC columns was 5 µm for a long time, until the mid-1990s, when 3.5 µm became popular for method development. More recently, as higher speed and/or higher resolution is required, chromatographers have turned to packings with sub-2-3 µm, including 1.8 µm particles made by Agilent. Shorter columns with these particles can produce faster high-resolution separations. The 3.5 µm particle size operates at a routine operating pressure and may be used on all LCs, including those with a 400 bar operating limit. Short (50 mm and shorter) 1.8 µm columns may be employed on optimized standard LCs, while longer columns may require a higher-pressure LC (e.g. Rapid Resolution LC or UHPLCs), operating at pressures from 600 to 1200 bar. Recently, newer technology (e.g. superficially porous particles) has been developed that enables performance similar to sub-3 µm columns, but generates lower backpressure, so it can be used with conventional HPLC instruments (see Superficially Porous Particles, p.16).

    If the particle size of a column is reduced by half, the plate number doubles (assuming column length remains the same). However, if particle size halves, column backpressure increases four times (~1/dp

    2). If column length doubles, the plate number and analysis time also double. As column length increases, backpressure increases linearly. For example, a 2.1 x 100 mm column packed with 3.5 µm particles generates about 12,000-14,000 theoretical plates, an efficiency that can provide adequate separation for many samples. By reducing the particle size from 3.5 µm to 1.8 µm, the efficiency of the same 2.1 x 100 mm column is doubled to about 24,000 theoretical plates. However, this column generates a backpressure that is four times greater than the pressure of the same size column filled with 3.5 µm particles.

    Very often, an efficiency of 24,000 plates is not required, so the column length can be halved to 50 mm, with an expected efficiency of 12,000 plates. The analysis time will be cut in half with this shorter column and the backpressure is only twice as great as the 100 mm column with 3.5 µm particles.

  • 21

    When you want to establish a routine method, consider reducing the column dimensions to the smallest available size for your analysis and instrument; smaller columns are often less expensive to buy and use less solvent. In some cases, if column diameter is reduced by half, sensitivity increases by four to five times (assuming the injection mass is kept constant). For example, when a sample is injected onto a 2.1 mm id column, the peaks are about three to five times higher than on an optimized LC than when the same amount of sample is injected onto a 4.6 mm id column. If your instrument is optimized for low-volume columns, as long as linear velocity is maintained, column efficiency, theoretical plates, backpressure and analysis time are not significantly affected by reducing the diameter of the column.

    Application objective Column diameter (mm)

    Very high sensitivity, LC/MS, peptides and proteins 0.1, 0.075

    Very high sensitivity, limited sample, LC/MS, peptides and proteins 0.3, 0.5

    High sensitivity, limited sample, LC/MS 1.0

    Save solvent; special low-volume instrumentation is available 2.1

    Special detectors such as MS 2.1

    High sensitivity, limited sample 2.1

    Save solvent; standard HPLC equipment available, LC/MS 3.0

    Standard separations 4.6

    Small-scale (mg) preparative separations 9.4

    Medium-scale preparative separations (100 mg to g) or semi-prep 21.2

    Large-lab-scale preparative separations (up to 100 g) 30, 50

    Pilot and process scale 100 mm to 1 m

    Column dimensionsFor many years the column sizes most often recommended for analytical method development were 4.6 x 150 mm or 4.6 x 100 mm with a 5 µm particle size. If more resolution was needed, a 4.6 x 250 mm column was recommended. But with the range of modern choices available analytical method development with 4.6 x 100 mm columns with 3.5 µm or 2.7 µm superficially porous particles are the recommended starting point.

    During method development, choose the column id (for example 2.1 or 3.0 mm) to accommodate additional application objectives (such as sensitivity, solvent usage) or compatibility with certain instrument types (capillary, nano, or prep columns).

    Nano, capillary or microbore columns are used when increased sensitivity is required or when the sample is extremely limited.

    • Nano columns for sample sizes below 1 pg used with nL/min flow rates

    • Capillary columns for sample sizes in the range pg to ng with flow rates around 4 µL/min

    • Microbore columns for sample sizes from ng to µg typically operate at flow rates around 40 µL/min

    Table 2. Applicationsandcolumndiameters

  • 22

    It is easy to calculate whether a shorter column will achieve the same results as your longer development column, by using data produced by the development column. Decision making by simple calculation will save you a great deal of time. You can use Agilent’s Method Translation Software (search for LCMethodTranslationSoftware at to help you with these calculations.

    Cartridge column systems

    Cartridge systems can provide flexibility and economy because the cartridge is fitted to existing column hardware. Cartridges for analytical systems are pre-packed, whereas you can pack your own systems for preparative and semi-preparative applications using media purchased in bulk.

    Type of cartridge Features Benefits

    Analytical Columns and Analytical/Guard combinations

    Agilent HPLC Cartridge Can reverse collets in the end fitting to add guard cartridges


    Extends column lifetime

    Permits rapid column changes

    Can use 2, 3, 4 and 4.6 mm cartridges

    Cartridges have a unique filter and sieve at each end

    Helps prevent blockage

    ChromSep: standalone system comprising holder, analytical cartridge, and guard column

    Wide combination of column lengths and diameters

    Modular flexibility

    Cartridges and guards in multi-packs Economical

    No need for special tools Easy to use

    ZORBAX Rapid Resolution and Rapid Resolution HT Cartridge Columns: 1.8 and 3.5 µm packings, standalone system

    For high throughput LC/MS, LC/MS/MS and combinatorial separations

    For all analyte types

    Packed with Eclipse XDB for pH 2-9 For all analyte types

    Packed with StableBond for low pH use Low bleed

    Sold individually or as three packs For all analyte types


    ZORBAX Guard Cartridge: standalone system

    High efficiency, standalone, low dead volume cartridge

    Seals up to 340 bar

    Polymeric cartridge designed for leak-tight seals against metal surfaces

    No gaskets required

    Reusable fittings Adapt for connections to 1/16 in. LC fittings

    ContinuedonnextpageTable 3. CartridgeSystems

  • 23

    Type of cartridge Features Benefits

    Semi-Preparative Guards

    ZORBAX Semi-Preparative Guard HPLC Hardware Kit: standalone system

    Easy low-dead-volume assembly Seals up to 2,000 psi (135 bar, 13.5 MPa)

    Tubing (polyphenylene sulfone) designed for leak-tight seals against metal surfaces

    No gaskets required

    Reusable fittings Adapt for connections to 1/16 in. LC fittings

    ZORBAX and Agilent Prep Cartridge Column and Guard HPLC System: standalone and integral hardware options

    Easy low-dead-volume assembly Extends column lifetime

    Reusable fittings Permits rapid column changes

    Hardware options for integral and external guards

    Use with 21.2 and 30 mm id columns


    Load & Lock Preparative Columns and Packing Station: standalone system for dynamic and static ‘locked’ axial compression

    For laboratory and process purification, at high quality and high volume with three column sizes, up to 24 in. id

    Easy scale-up from g to multi-kg quantities

    Mobile packing station Use anywhere

    Runs on compressed air Safe to use in hazardous environments

    Quick release to pack and unpack in minutes Maximize productivity

    Dynamax Preparative Cartridge Column and Guard: standalone dynamic axial compression system

    Modular design with reusable end fittings Reduced hardware costs

    10, 21.4 and 41.4 mm id Easy scale-up

    Integral guard column option Delivers longer column lifetimes with complex samples


    Polymeric PLRP-S, PL-SAX and PL-SCX

    Range of pore sizes and particle sizes to give high sample throughput

    Increased productivity

    Chemical and thermal stability enables cleaning in place and sanitization

    Increased column lifetime

    Robust packing methods for process hardware

    Improved packed column performance

  • 24

    Column type Guard cartridge holder ID (mm) Phases

    Cartridge/Guard Cartridge Systems Compatibility Guide*

    Cartridge column cartridge holder 5021-1845

    Guard cartridge (internal system) cartridge holder 5021-1845











    Standard fitting Column guard cartridge (standalone) cartridge holder 820888-901





    Rapid Resolution cartridge holder 820555-901

    No guard cartridge holder 4.6 ZORBAX

    Semi-preparative column Semi-prep guard cartridge (standalone) cartridge holder 840140-901

    9.4 ZORBAX

    PrepHT (no photo available) Guard cartridge 820444-901

    21.2 ZORBAX

    Agilent Prep

    *Standalone guard cartridges fit all cartridge and standard fitting columns available from Agilent.

    Table 4. Cartridge/GuardCartridgeSystemsCompatibilityGuide

  • 25

    Keys to performance: column configurations and settings

    There have been many books written about chromatography, and there are many more 'keys to performance' than we can possibly cover here.

    Our objective here is to highlight a few areas which are often overlooked or tend to cause confusion.

    We start with system and 'mechanical' items:

    • Reducing extra-column volume

    • Making good fittings

    • Sample injections

    • Understanding and measuring system dwell volume

    • Setting the data collection rate for high efficiency columns

    Next, we move to topics that have more to do with the process of chromatography, and method development:

    • Understanding chelating compounds

    • Evaluating pH

    • Working with gradients

    • Optimizing column re-equilibration

    We finish up the section with a few things you'll likely need to deal with over time:

    • Column aging

    • Column cleaning – for reversed-phase (silica-based vs. polymeric) and normal phase

  • 26

    The importance of reducing extra-column volume

    Extra-column volume refers to the 'extra' or 'external' (to the column) volume that is part of your system, most specifically, the connecting tubing carrying the sample aliquot between your LC components and your column, your injection volume and your flow cell volume.

    The larger your column, the larger your column volume (Vm), and the less important it is to reduce your extra-column volume. However, when you use smaller, high efficiency columns such as Poroshell 120 and sub-2 µm particle columns, you’ll want to reduce your extra-column volume as much as possible, to avoid seeing it impact your chromatographic results. Unnecessary extra-column volumes may lead to a loss of efficiency and, in some instances, tailing.

    The following examples show the effects of extra-column volume (Figure 9). A small 4.6 x 30 mm column with only 10 µL of extra-column volume produced a nice separation. The same column with added extra tubing, which created an extra-column volume of 50 µL, gave broader peaks and loss of resolution as can be seen by comparing the first three peaks in both chromatograms. This situation is often created when someone introduces tubing to a system without knowing the volume. The addition of a piece of tubing that is too long, or a short tube with a larger id, can lead to poorer chromatography for very efficient columns. It is important to know the internal diameter tubing you are using and what volume it adds. Also, if you are using smaller-volume columns, you may need to use a micro or semi-micro flow cell. Standard flow cells increase extra-column volume.

    As a guideline, the maximum extra column volume of the system (tubing volume, injection volume and detector volume) should not exceed 10% of the column volume.

    Agilent sells capillary kits which contain capillary tubing and Swagelok connectors in various sizes, so you can find just the right length to minimize your connections and your tubing volume. Tubing is color-coded, to identify the diameter of the tubing. As you move to high efficiency columns, you’ll want to use the narrow-diameter red tubing (0.12 mm id) for your connections, instead of the green 0.17 mm id tubing that is often used on conventional HPLC instruments.

    Peak identification1. Phenylalanine2. 5-Benzyl-3,6-dioxo-2-piperazine acetic

    acid3. Asp-phe4. Aspartame

    ConditionsColumn: StableBond SB-C18,

    4.6 x 30 mm, 3.5 µm

    Mobile phase: 85% H2O with 0.1% TFA:15% ACN

    Flow rate: 1.0 mL/min

    Temperature: 35 °C

    Figure 7. Theinfluenceoftubingvolumeonchromatographicperformance

  • 27

    Preparing the perfect fitting connection

    Problems with improper stainless steel tubing connections are often mistaken for column issues, and are the source of many calls to Agilent's technical support line.

    Connection issues can arise because different manufacturers supply different types of fittings (see Figure 10).

    You should ideally use the fittings that are recommended by your column manufacturer. Most analytical reversed-phase columns are compatible with Swagelok or Parker type fittings when seated correctly in the column that you will be using.

    Stainless steel fittings are the best choice for permanent high pressure sealing. Agilent recommends Swagelok type fittings with front and back ferrules that give the best performance throughout the Agilent LC system. You can use them on all instrument connections such as valves, heaters, column connections, etc. Alternately, for convenience and lower pressure operation to 600 bar, finger-tight polymeric fittings allow for adjustment of the end-fitting to seat the capillary into the column properly, helping to avoid extra-column voids and leaks. These connectors can be tightened without wrenches. There are also new high pressure fittings, for use up to 1200 bar (PN 5067-4733) which are designed to be able to be removed and resealed. See Figure 9 for examples of the various fittings.

    Figure 8. ColumnconnectorsusedinHPLC

  • 28

    PEEK and polyketone fittings are ideal for connections that need to be changed frequently, or when biocompatibility is required. They are also useful where pressure is less critical, such as the exit fitting on a column. PEEK fittings can be used up to 200 bar and polyketone fittings up to 600 bar.

    Making a good connectionRelative to the distance from the end of the tubing to the bottom of the ferrule, tubing can be too long or too short – these situations can lead to leaks or peak tailing/splitting, as shown in the illustration on p 29. If the tubing is too long, the ferrule will not seat properly and leaks will occur (see Figure 10). If the tubing is not pushed in far enough, a void space is created that acts as a mixing chamber and introduces extra-column volume, resulting in poor peak shape. If you use columns from different manufacturers make sure you use the correct fittings and that the fitting is correctly seated in the column end fitting.

    Which type is used when?

    • Agilent uses Swagelok pressure-type fittings with front and back ferrules - which gives best sealing performance - throughout all our LC systems

    Stainless steel (SS) fittings are the best choice for reliable high pressure sealing

    • Connections are changed frequently, i.e. connecting columns

    • Pressure is less critical

    PEEK (

  • 29

    Steps for making good connections

    Step 1:Select a nut that is long enough for the fitting you’ll be using

    Step 2:Slide the nut over the end of the tubing

    Step 3:Carefully slide the ferrule components on after the nut and then finger-tighten the assembly while ensuring that the tubing is completely seated in the bottom of the end fitting.

    Step 4:Use a wrench to gently tighten the fitting, which forces the ferrule to seat onto the tubing. Don’t over-tighten it, though, because that will shorten the useful life of the fitting.

    Step 5:Once you believe you have the fitting complete, loosen the nut, and inspect the ferrule for the correct position on the tubing

    Step 1

    Step 5 - a perfect fitting

    Step 3

    Step 2

    Step 4

    Figure 10. Examplesofincorrectfittings

    Figure 11. Stepstomakeaproperfitting

  • 30

    Sample injection

    The injection volume of your sample is important to your results. If you have your injection volume too large the column can be overloaded, which will lead to peak broadening, most often peak fronting or in some cases, peak tailing.

    In Figure 12, we see an injection using 1, 2, 5 and 10 µL injections, and you can see how the peaks broaden as the injection volume is increased.

    Figure 12. Sampleloadingvolumescompared

    To see more tips for making connections in action, check out the LC Troubleshooting series for more information:, and look for the Peak Broadening Video.

    After you have decided your injection volume, you need to make sure your injection solvent is similar enough to the mobile phase to reduce band broadening or splitting.

    ConditionsColumn: ZORBAX SB-C18,

    3.0 x 50 mm, 1.8 µm

    Flow rate: 1 mL/min.

    Temperature: 45°C

    Mobile phase: 30 - 65% ACN at 2.5 min.

    Detection: 250 nm UV

    Diluent: methanol

    Sample: methyl, propyl and butyl paraben

  • 31

    The sample injection solvent and volume can have an impact on peak shape. In the example (Figure 13), even a 5 µL injection of a paraben sample in a methanol solvent shows evidence of the beginning of peak band broadening. The reversed-phase column had the dimensions of 3 x 50 mm and the instrument was operated in a low dispersion configuration. With a 10 µL injection, the loss of peak symmetry is clearly a problem. In reversed-phase LC, 100% organic, or 100% of the strong solvent (in this case methanol), for larger injection volumes, will cause the peaks to be prematurely swept down the column, resulting in peak distortion. The problem can be overcome by incorporating an evaporation step to concentrate the analytes so that a smaller injection volume can be used. Alternatively, the injection solvent can be diluted with water to make it more compatible with the mobile phase, allowing for larger injection volumes without peak distortion.

    Figure 13. Solventeffects:strongdiluentswith1-10μLofinjectionvolume

    Setting the data collection rate

    When using small volume columns, the data collection rate is a common source of ‘artificial’ peak broadening. For rapidly eluting peaks, you’ll want to make sure that you sample enough points across the peak so that the algorithms in the data system can accurately determine peak widths, peak area and retention time. If you take too few points by sampling slowly (e.g. low data collection rate), your peaks appear wider than they actually are as they elute from the column. You’ll want to look at your data collection rate and ensure it is properly set to optimize your results for the specific column you are using. Figure 14 illustrates the impact of the data collection rate.

    If you are using the Agilent Method Translator with your current column and gradient conditions, the translator will estimate the expected peak width (5 sigma) under your conditions. That information may be helpful when setting up the initial method.

    ConditionsColumn: ZORBAX SB-C18,

    3.0 x 50 mm, 1.8 µm

    Flow rate: 1 mL/min.

    Temperature: 45°C

    Mobile phase: 30 - 65% ACN at 2.5 min.

    Detection: 250 nm UV

    Sample: methyl, propyl and butyl paraben

    Diluent: methanol

  • 32

    Here, we’re measuring the efficiency, and as you can see, efficiency increases as the data collection rate used went higher.

    You can optimize your data collection rate by adjusting the detector setting and/or the time constant to the fastest possible value that does not compromise signal-to-noise. The peak width control in ChemStation enables you to select the peak width, or response time, for your analysis. The peak width, as defined in the ChemStation software, is the width of a peak at half height. Set the peak width to the narrowest expected peak in your sample. You should not use a faster response time than you need since this may lead to greater noise at the baseline.

    Dwell volume and its impact on chromatography

    For low pressure mixing systems dwell volume equals all of the volume from the proportioning valve, through the pump and other system components and on to the head of the column (see Figure 15). For high pressure mixing systems, dwell volume equals all of the volume from where the solvents first meet, after the two metering pumps, to the head of the column (see Figure 16).

    Figure 14. ComparisonofpeakefficiencyonPoroshell120EC-C184.6x100mmwithdifferentdatacollectionrates

    ConditionsColumn: Poroshell 120 EC-C18,

    4.6 x 100mm

    Instrument: 1200 SL 2 µL flow cell

    Flow rate: 2.00 mL/min

    Sample: 2 µL injection.

    Mobile phase: 60:40 MeCN:Water

    Figure 15. Dwellvolume:lowpressuremixingquaternarypump

  • 33

    For gradient separations, dwell volume imposes a de facto isocratic hold time at the beginning of the gradient, equal to the dwell volume divided by the flow rate. Too large a dwell volume makes some instruments impractical or unusable for narrow-bore gradient separations.

    When using narrow-bore columns, instrument configuration is crucial. Both dwell volume and extra-column volume must be minimized for optimal use of narrow-bore (2.1 mm id) and microbore (1 mm, and < 1 mm id) columns.

    Figure 17 shows a chromatographic example that demonstrates the effect of dwell volume on analytical results. Note that the early eluting peaks are broader. The issue here is the early peaks are eluting quite late due to the dwell volume - essentially, they are eluting isocratically. If this interferes with separation or detection, we will need to reduce the dwell volume or move the application to another system.

    Figure 16. Dwellvolume:highpressuremixingbinarypump

    Peak identification1. phenacetin2. tolmetin3. ketoprofen4. fenoprofen

    5. ibuprofen6. phenylbutazone7. mefenamic acid8. flufenamic acid

    Figure 17. Impactofdwellvolumeongradientchromatographicresults

    ConditionsColumn: ZORBAX Rapid

    Resolution Eclipse XDB-C8

    Mobile phase: A: 0.1% TFA in H2O B: 0.1% TFA in ACN

    Temperature: 35°C

    Gradient: Case 1: 15-60% B/ 6 min, 1.0 mL/min Case 2: 30-75% B/6 min, 0.2 mL/min

  • 34

    Using a ruler on paper or by inserting lines in PowerPoint, draw lines parallel to the x or y axis at the following points:

    1. At the zero signal defined by the retention of 90% B

    2. At the maximum stable signal defined by the region 100% B

    3. Vertical line at 2.0 minutes. Your image should look like Figure 19.

    Figure 19. Calculatingdwell(delay)volumes,steptwo

    Measuring your system’s dwell volumeStart by replacing the column with a short piece of HPLC stainless steel tubing. Prepare mobile phase components as follows: A—Water (UV-transparent), B – Water with 0.2% acetone (UV-absorbing). Monitor at 265 nm. Run the gradient profile at 0 - 100% B/10 min. at 1.0 mL/min. Record, then print out your gradient trace (Figure 18).

    Figure 18. Calculatingdwell(delay)volumes,stepone

  • 35

    Follow the x axis to the 50% B vertical line and determine the time, as closely as possible, at which the 50% response is observed. An error of 0.1 minute will be a 50 µL error in your delay volume estimate, assuming your test was run as we did at 500 µL/min flow rate. In this example, we estimated the 50% time at 4.04 minutes.

    To calculate the dwell volume the simple formula is Time (50%) - Time (step) x Flow Rate = Dwell Volume. In this example, (4.04 - 2.00) minutes x 500 µL/min = 1020 µL.

    This measurement was made on an Agilent 1200 RRLC system (1200 SL) with nominal 340 bar backpressure via a nominal 100 cm PEEK restrictor with 0.062 mm (0.0025 inch) internal diameter using water as the specified solvent for compressibility compensation in flow delivery. The system included the standard mixer and pulse damper and the autosampler was used in the normal flow path (vs. bypass mode) to a restrictor as described and fitted with a 3 mm 2 µL flow cell.

    After the three lines have been added, add two additional lines. Calculate the 50% response of the step, in this case, 24.5 mAU, and draw a horizontal line across the image. Next, draw a vertical line that intersects the junction of the 50% response line and the observed detector signal. Your new image should look like Figure 20.

    Figure 20. Calculatingdwell(delay)volumes,stepthree

  • 36

    Here, the systems with the smallest and largest dwell volumes did not produce good results. It is likely that this method was originally developed on an instrument with a dwell volume in the medium range.

    To investigate the effect of dwell volume on your gradient separation, you can either increase or decrease the isocratic hold at the beginning of the gradient run.

    To simulate larger dwell volume on a system with lower dwell volume, set a hold time at the beginning of the gradient program, equal to the difference between the two dwell volumes, in mL, divided by the flow rate in mL/min.

    To simulate smaller dwell volumes than that of your instrument, you must modify the injector program so you impose an injection delay time equal to the same dwell time. Some instruments have this capability; others do not.

    If the method is to be transferred to another laboratory with a different instrument, it is extremely important to document, in your written method, the dwell volume of the instrument(s) used in method development, and any effect of dwell volume on the separation.

    Figure 21. Dwellvolumeandresolution

    Evaluating the impact of dwell volumeRunning a method on instruments with different dwell volumes can produce erratic results, as demonstrated in Figure 21.

  • 37

    Dwell volume and analysis timeDwell volume also includes injector volume, so it is important to minimize the internal volume and connecting tubing around the injector. Most Agilent autosamplers offer the option of running in the bypass mode or with Automatic Delay Volume Reduction (ADVR), where the injector is switched back to the load position after the sample is flushed from the sample loop. Depending on which injector you are using, this reduces volume in the system by up to 300 µL. It is possible to start with a delay volume of about 1100 µL on a standard Agilent 1200SL or 1260 Infinity binary gradient instrument and get it down to about 280 µL, by changing to smaller diameter tubing, using the autosampler bypass function, and removing the mixer and damper.

    You can reduce analysis time in rapid gradient separations by overlapping injections. In an overlapped injection, the autosampler draws up the sample during the previous run and injects when the system is ready. This can reduce run time by about 30 seconds or more per run. If you are doing fast, gradient combinatorial chemistry runs, this is up to 1/3 of your analysis time.

    If you are going to do a lot of gradient work with very small columns, like the capillary and microbore columns, a delay volume of 280 µL will be too high. It is important to choose the Agilent Capillary LC instrument for those columns.

    Don’t forget that it is also important to minimize extra column volume and inject in an appropriate sample diluent. This helps minimize band broadening that degrades the resolution.

    In addition, as discussed earlier, it is necessary to make sure the detector response time is set correctly to capture enough points on the quickly eluting peaks. If it is not, then distorted, artificially broad peaks may appear.

    Chelating compounds

    Some analytes have structures that are suitable for chelating metals. Metal on the frit or on the column walls may interact with a chelating compound. A phosphoric acid wash may help to eliminate this problem.

    Here is a good way to tell if metals pose a problem. With a compound that has a lone pair of electrons, the addition of a metal may form a ring structure; this may cause irreproducible retention or a peak shape issues. (Figure 22).

    Figure 22. Evaluatingcompoundstoidentifychelatingissues

  • 38

    In a case such as this, a phosphoric acid wash may be beneficial. Notice in the ‘before’ chromatogram, compound 2 has a high peak-tailing factor (Figure 23). After the acid wash, we see a sharper peak and an improved tailing factor on compound 2. A 1% phosphoric acid wash is acceptable for ZORBAX StableBond columns because they are designed to be rugged in acidic conditions. If you use an Eclipse-XDB, Eclipse Plus, or any end-capped column that is designed for the mid pH range, reduce the acid concentration to 0.5%.

    Figure 23. Washingwithphosphoricacidrestorescolumnperformanceforchelatingcompounds

    ConditionsColumn: ZORBAX SB-Phenyl,

    4.6 x 150 mm

    Sample size: 5 µL

    Mobile phase: 75% 25 mM ammonium phosphate

    Buffer 25% ACN

    Flow rate: 1.0 mL/min

    Temperature: Ambient

    pH and mobile phase modifiers

    The pH of your mobile phase and your sample can have a big impact on your method development (see ‘Working with Mobile Phases’ and ‘Method Development’ sections for more information). If possible, it is good to avoid extremes – either high or low – in pH, as working at these extremes may shorten the lifetime of your column.

    Review the recommended pH range of the column you are using for the optimal pH range. Most separations will take place between pH 2 and 8. The retention of acidic and basic – ionizable – analytes will be sensitive to pH and can often change dramatically with pH. Evaluating the optimum pH for your separation is a key part of method development and more detail can be found in the Method Development section.

  • 39

    Working with gradients

    The more complex your sample, the more likely you are to use a gradient method. Gradient separations are useful for compounds that differ widely in polarity, or are of high molecular weight, such as peptides and proteins.

    Although isocratic methods are easy to use, peak broadening of late-eluting peaks can occur because peak width increases with retention time. Gradient elution can overcome this problem by decreasing the retention of late-eluting peaks. Benefits of gradients include sharper peaks because of gradient compression effects and reduced build-up of contamination, because the gradient exposes the column to continuously increasing solvent strength.

    Figure 24 shows a comparison between an isocratic and gradient method. When separated isocratically, this eight-component herbicide mixture is not fully resolved in 70 minutes, and components 1 and 2 co-elute. Reducing the percent organic to resolve peaks 1 and 2 will only lead to an unacceptable retention time and possibly unacceptable limits of detection for peak 8. When using a gradient of 20-60%, all eight components are well-resolved in less than 30 minutes with essentially equivalent limits of detection. The analysis time could be further reduced by starting the gradient at a higher percent organic. One way to do this would be to run a 25-65% organic.

    Gradient separation for reversed-phase chromatography is accomplished using two to four mobile phase components. In a binary gradient they are usually referred to as A and B. The A solvent is weaker (often water or buffered water), and allows the analyte to slowly elute from the column. The B solvent is stronger and causes the analyte to elute more rapidly. B is an organic solvent miscible with water, such as acetonitrile (ACN), methanol (MeOH), tetrahydrofuran (THF), or isopropanol (IPA). Start by running a test during gradient development.

    Here are the key steps to creating a gradient:

    1. Start by running a test using a linear gradient from 5 - 10% to 100% organic over a set time period. Limit your gradient to ~70% organic if your buffer is insoluble in the B solvent.

    2. Hold the final composition for some additional time to ensure all sample components have eluted.

    3. Examine the chromatogram to determine the appropriate initial gradient composition and gradient profile.

    Most gradient separations are achieved using a linear gradient with a constant change in organic composition over time. However, other types of gradient are possible. For example, some gradient programs involve different slopes at different times, or sudden step changes in the relative concentrations of A and B, depending on the separation and retention time needed.

    See more about gradients, and some examples, in the Method Development section.

    Keep your gradient simple (using shorter gradient times) and use a shorter column (50-75 mm) to reduce run times and improve peak detection.

  • 40

    Optimizing column re-equilibration

    Sufficient column equilibration is necessary so your retention times during method development or method transfer are reproducible. Looking at the gradient separation in this grapefruit juice analysis (Figure 25), we can see the column void volume and amount of time dedicated to column re-equilibration.

    ConditionsLC: Agilent 1290 Infinity

    with DAD

    Column: Agilent RRHD SB-C18 2.1 x 150 mm, 1.8 µm PN 859700-902

    Injection : 1 µL of 0.2 µm filtered grapefruit juice

    Temperature: 40 °C

    Detection: UV, 276 nm

    Solvent A: Water

    Solvent B: AcetonitrileFigure 25. Gradientexample:grapefruitjuiceanalysis

    Figure 24. Comparisonofisocraticandgradientanalyses.

    Peak Identification1. Tebuthiuron2. Prometon3. Prometryne4. Atrazine5. Bentazon6. Propazine7. Propanil8. Metolachlor

    ConditionsColumn: ZORBAX SB-C8

    4.6 x 150 mm, 5 µm

    Mobile phase: A: H2O with 0.1% TFA, pH 2

    B: Acetonitrile

    Flow rate: 1.0 mL/min

    Temperature: 35 °C

    This is the time it takes to return the column to starting mobile phase conditions after a previous gradient has run. In this case, re-equilibration time was 1.4 min, equivalent to 4 column volumes, for this column size.

  • 41

    Why does this have to be done by trial and error?

    • The rate of stationary phase re-equilibration back to the starting conditions varies. How slow or fast you go depends on the mobile phase solvents, the buffers and gradient range used. This is why ten column void volumes are recommended for method development, to be sure the stationary phase is re-equilibrated for the next experimental run. Once you have the separation you want, you can then look at shortening the re-equilibration time.

    • Because the equilibration volume includes the dwell volume of the instrument, this additional volume also can vary from instrument to instrument. Dwell volume has no effect on retention in isocratic methods, but in gradient methods it has a significant effect on retention. If you don't know your system's dwell volume, it could have serious implications, because it's possible that your dwell volume is much larger than your column volume. If you move a gradient method from one instrument to another and the retention times shift, compare the dwell volumes of the two systems. The closer the flow path volumes match, the closer the retention times will match.

    Tip: View the pressure trace in the online signal window. The pressure changes as the mobile phase composition changes. When the pressure trace returns to its starting value, you know you are near re-equilibration.

    ConditionsLC: Agilent 1290 Infinity

    with DAD

    Column: Agilent RRHD SB-C18 2.1 x 150 mm, 1.8 µm PN 959759-902

    Injection: 1 µL of 0.2 µm filtered grapefruit juice

    Temperature: 40 °C

    Detection: UV, 276 nm

    Solvent A: Water

    Solvent B: Acetonitrile

    Sample: Grapefruit juiceFigure 26. Overlayofchromatogramsconfirmsequilibration

    You can estimate your column void volume by looking at a chromatogram and measuring the time from the start of the chromatogram to the first disturbance on the baseline. Take this time in minutes and divide by the flow rate in mL/min to find the mL of void volume. Alternately, use the Agilent Method Translator which will calculate the void volume for you.

    The rule of thumb is to use ten column volumes of equilibration during method development, but for established methods, the equilibration time can be evaluated and shortened, for instance, like in this case to 4 column volumes. The minimum equilibration time has to be determined by trial and error, by running the gradient repeatedly and looking for any changes in retention. If several repeated injections are identical, as in Figure 26, equilibration time is sufficient.

  • 42

    Column aging

    As you use your column and expose it to various mobile phases, analytes, and sample matrices, it will be subtly affected by these interactions and over time, there may be changes in its resolution. It’s important to keep a test chromatogram, from when your column was new, to compare and understand these changes in performance over time. If resolution degrades beyond an acceptable value for good quantitation, the column should be discarded and replaced with a new one.

    Loss of bonded phaseWhile column contamination is one of the most common reasons for decline in silica-based column performance, loss of bonded phase will change retention over time. If certain non-compatible solvents are used in the mobile phase, the bonded phase can be stripped away from the base silica, leading to changes in resolution and retention times, and an increase in peak tailing. Loss of bonded phase is most severe below pH 2. You can limit its effects by using a stabilized bonded phase column, or by operating between pH 2 and pH 7, and below 60 °C. On the high pH side, the dissolution of the underlying silica gel can lead to a loss of packing, formation of a void and poor peak shapes. If you must work at pH values greater than 9, make sure you use a specialty bonded phase column suited for this operation or a polymer-based column which can easily withstand high pH mobile phases.

    To determine when a column is wearing out, you should monitor and look for changes in its capacity factor, selectivity and tailing factor. This will also help you predict when your column needs to be replaced.

    Cleaning a reversed-phase silica column

    Sometimes you can alleviate a high pressure or peak tailing issue by cleaning your column. Before cleaning your column, we recommend that you disconnect it from your detector and run your wash solvents into a beaker or other container. In some cases, it is advisable to backflush the column while cleaning it. This keeps the column contamination from flowing through the column. Check with your column manufacturer, or review the instructions that came with your column before doing this.

    Cleaning the column requires solvents that are stronger than your mobile phase. You should use at least ten column volumes (see Figure 27) of each solvent for cleaning analytical columns. See the following procedure for cleaning reversed-phase columns. When finished cleaning your column, return your column to the flow direction recommended by the manufacturer.

    Tip: Know your system's dwell volume! Check out p. 34 for a method to do this. Check your system's documentation, or contact Agilent tech support for help. See Agilent resources section in this book.

    A pressure problem may not be due to column blockage, but to seals breaking down in your system and depositing particulates on the frit in your inline filter; try changing your inline filter frit before cleaning your column.

  • 43

    Tip: To avoid precipitation if there are non-volatile buffers in your system, first flush with non-buffered aqueous mobile phase before introducing pure organic solvent.

    Figure 27. Calculatingyourcolumnvolume

    Do not attempt to change the inlet frit on your column. Due to the high efficiency processes used today to pack columns, this may cause irreversible damage to your column

    Steps to backflush or clean your column:

    1. Disconnect the column from the detector, attach tubing to the end of the column and place it in a beaker to capture the liquid.

    2. Start with your mobile phase without buffer salts (water/organic)

    3. Next, use 100% organic (methanol and acetonitrile)

    4. Check pressure to see if it has returned to normal; if not, then

    5. Discard column or consider stronger conditions: 75% acetonitrile/25% isopropanol

    6. 100% isopropanol

    7. 100% methylene chloride*

    8. 100% hexane*

    *When using either hexane or methylene chloride, the column must be flushed with isopropanol prior to use and before returning to your reversed-phase mobile phase.

  • 44

    Buffer salts can precipitate out and cause backpressure build-up inside the column. If this occurs, slowly pump warm water through the column to remove them.

    Wash solutions containing propanol will generate higher operating pressures due to increased viscosity. You may need to reduce the flow rate during that stage of cleaning to maintain safe pressure operation.

    Cleaning a normal phase silica column

    For normal phase, you can only use organic solvents. Use at least 50 mL of each solvent, assuming your column is the traditional analytical 4.6 x 250 mm column (20 column volumes). Try these solvents, in order of increasing strength:

    1. 50% methanol:50% chloroform

    2. 100% ethyl acetate

    Cleaning a column used in normal phase mode may depend on the sample type. If these solvents do not work, contact Agilent so we can recommend a solvent that may be more effective with your application and sample matrix.

    Cleaning a reversed-phase polymeric column

    We strongly recommend that you maintain a minimum of 1% organic modifier in the mobile phase when using polymeric reversed-phase columns. Column performance can be degraded when reintroducing organic mobile phases after prolonged use in 100% aqueous eluents; nor should columns be ‘washed’ with 100% aqueous buffer. You can reverse the direction of flow, and if starting pressures are high, reduce the flow rate. Here are the steps to clean a reversed-phase polymeric column:

    1. Run a clean-up gradient using the current mobile phase, moving to a high-percent organic (95%) and hold for several column volumes. Repeat this step two or three times. Buffers may precipitate at high organic. If your mobile phase contains a buffer, you may want to limit the organic to ~70% or flush out the buffer with the mobile phase minus the buffer prior to cleaning.

    2. Try a higher-strength organic modifier such as TFA to remove hydrophobically-bound contaminants

    3. Peptide or protein contamination may sometimes be removed by aqueous/ACN containing 1.0% v/v TFA

    4. Strong acids and bases, including 1 M sodium hydroxide, can be used for cleaning in place and depyrogenation

    Tip: Filtering your sample will increase the lifetime of your column and will help reduce instrument wear.

  • 45

    Method development

    If you don't have an established HPLC method, you'll need to do some method development in order to have a robust and reproducible analysis. Having the skill to develop a successful method will help you when encountering difficult assays in the future.

    In this section, we start off broadly with an overview that encompasses all HPLC modes:

    • Key steps of method development

    • Mode selection

    • Column packing and dimensions

    • Stationary phase selection

    Then, we focus on reversed-phase method development:

    • Selection of stationary phase for reversed-phase chromatography

    • Selection of mobile phase for reversed-phase chromatography

    • Tips for working with mobile phases

    • Troubleshooting examples involving mobile phase and mobile phase modifiers

    • Managing pH with mobile phase modifiers

    • Troubleshooting examples involving pH

    • Reversed-phase method development with polymeric columns

    • Tips for transferring existing methods on conventional columns to superficially porous particle columns

    • A step-by-step guide to 'hands-on' method development

    • Automated method development

  • 46

    At the end of this section, we also discuss method development for other HPLC modes:

    • HILIC

    • Normal Phase Chromatography

    • Ion-exchange Chromatography

    • Gel Permeation/Size Exclusion Chromatography

    Method development: where to start

    The overall goal in method development is to optimize the resolution for the desired analyte(s) in the shortest possible time. When you start out thinking about method development, it's a good idea to review the key drivers of resolution. So, as a refresher, you may want to review the Essential Chromatography Concepts section (pages 5 - 11)

    A typical method development scheme has the following steps, which will guide our content in this section:

    1. Choose the mode

    2. Choose the column and column packing dimensions

    3. Choose the stationary phase chemistry

    4. Choose the mobile phase solvents

    5. If the mode requires it, adjust the mobile phase pH

    6. Run some initial isocratic or gradient experiments to define boundary conditions

    7. Optimize the experimental conditions

    Mode SelectionThe first thing you should do is to choose the HPLC mode. The mode is generally decided by the type and solubility of the analyte(s) of interest, its molecular weight (MW), the sample matrix and the availability of the appropriate stationary phase and column. Figures 28 and 29 outline the steps for evaluating the best choice of mode, based on the molecular weight of your compound and the solvents you’ll be using.

  • 47

    Solvent Mode


    Molecular weight 2,000

    Gel permeation

    Gel filtration

    Ion-exchange with wide-pore material

    Reversed-phase with wide-pore material

    Figure 28. LCandLC/MScolumnselectionbysolventandmode

  • 48

    Molecular size

    Solvent Compound class Separation mechanism

    Small Aqueous/Organic

    Lipids Silver ion complexation

    Polycyclic aromatic hydrocarbons Reversed-phase C18

    Organic acids Ligand interaction reversed-phase ion pair

    Monosaccharides and disaccharides Ligand interaction

    Normal phase amino


    Oligosaccharides Ligand interaction and ion-exchange

    Sugar alcohols Ligand interaction

    Normal phase Normal phase amino/cyano/diol

    Basic, polar Polar reversed-phase C8 or C18 or HILIC

    Reversed-phase C18 ion suppression

    H-bonding Reversed-phase C8 or C18

    Positional isomers Reversed-phase C8

    Aromatic or structurally similar Reversed-phase phenyl/phenyl-hexyl/diphenyl

    Very polar Reversed-phase other or HILIC

    Extreme conditions Reversed-phase polymeric

    Organic Non-polar Normal phase Si

    Polar Normal phase amino/cyano/diol

    Sometimes, more than one mode may work for a particular set of analytes. For example, ionic compounds can be separated by ion-exchange chromatography on a resin or silica-based column or on a reversed-phase column using ion pair partition chromatography.

    Many chromatographers start with reversed-phase HPLC since there are many published applications. Reversed-phase chromatography can be used for non-polar, nonionic, ionic, and polar compounds and with a judicious choice of mobile phase and operating conditions, sometimes the entire analysis can be accomplished by this mode alone. We'll discuss the other modes at the end of this section, after we cover reversed-phase.

    Figure 29. LCandLC/MScolumnselectionbyanalyteandseparationmechanism

  • 49

    Choosing the column and packing dimensionsFigure 30 shows some of the parameters to consider when evaluating a column stationary phase and column dimensions. To perform high throughput analysis, a short column with small particles (e.g., sub-2 µm) may be the best choice. If you have a complex separation involving many sample components, then a long column packed with small particles could be chosen, keeping in mind that the operating pressure of such a column may increase dramatically. If you are performing mass spectrometry, a small internal diameter column (e.g. 2.1 mm id) may be the best choice, due to the lower flow rates used with an MS detector. For preparative chromatography, larger particles (5 or 10 µm) packed into larger diameter columns are often used. For such columns, it is preferable to have a higher flow rate pump to match the flow requirements of a preparative column.

    The pore size of the packing is important since the molecules must 'fit' into the porous structure in order to interact with the stationary phase. Smaller pore size packings (pore size 80 to 120Å) are best for small molecules with molecular weights up to a molecular weight of 2000. For larger molecules with MW over 2000, wider pore packings are required; for example, a popular pore size for proteins is 300Å.

    For most separations, stainless steel column hardware is sufficient. However, if you are analyzing fragile molecules that may interact with the metal surface such as certain types of biomolecules, then column materials such as PEEK or glass-lined stainless steel might be used. For the separation of trace cations, sometimes PEEK columns are the most inert. Note, though, that PEEK columns are limited to 400 bar.

    HPLC column

    Stationary phase Column dimensions

    Chemical propertiesChemical lifetime/Sensitivity

    Retention Factor

    Type of surface

    Physical propertiesEfficiency


    Pore size LengthInner

    diameterParticle size

    Figure 30. Somecolumnandchemistryeffects

  • 50

    Choosing the stationary phaseThere are a wide variety of stationary phases that are available for each of the modes. Many chromatographers practicing reversed-phase chromatography start with the most popular phase, octadecylsilane (C18), especially for small molecule separations. We will focus on reversed-phase separations in the next discussion but will cover other modes later in the chapter.

    Method development for reversed-phase chromatography

    Reversed-phase chromatography is by far the most common type of method used in HPLC - it probably accounts for 60% of all methods, and is used by nearly 95% of all chromatographers.

    In reversed-phase chromatography, we partition analytes between the polar mobile phase and the non-polar stationary phase – the opposite of normal phase chromatography. Typically, we get non-polar, non-specific interaction of analytes with hydrophobic stationary phase, meaning the sample partitions into the stationary phase. We use stationary phases like C18, C8, phenyl, or C3, which give polarity discrimination and/or discrimination based on the aromatic structure of a molecule.

    More polar analytes are less retained than non-polar analytes in reversed-phase chromatography. Retention is roughly proportional to the hydrophobicity of the analytes. Those analytes that have large hydrophobic groups and with longer alkyl chains will be more retained than molecules that have polar groups (e.g., amine, hydroxyl) in their structure. If you have a series of fatty acids, such as C12, C14, C16 and C18, the C12 would be the least retained and the C18 would be the most retained.

    The mobile phase is comprised of two main parts:

    1. Water with an optional buffer, or perhaps an acid or base to adjust pH

    2. Water-miscible organic solvent.

    Reversed-phase chromatography is quite versatile and it can be used to separate non-polar, polar, ionizable and ionic molecules, sometimes in the same chromatogram. Typically, with ionizable compounds, to improve retention and peak shape, we will add a modifier to the mobile phase to control pH and retention.

    Selection of stationary phase for reversed-phase chromatography

    Let’s consider an approach to developing a reversed-phase chromatography method. Figure 31 gives a general flow chart on how to select an appropriate stationary phase based on the molecular weight of a particular analyte. First, the pore size must be chosen to ensure that the molecules of interest will penetrate the packing material and interact with the hydrophobic stationary phase within the pores. Next, we choose the stationary phase; most start with a C18 phase initially. Depending on the ultimate goal of your method, you may choose a conventional analytical column, or if you are interested in high throughput, you might choose a ‘fast analysis’ reversed-phase chromatography column.

  • 51

    Most chromatographers begin with a C18 stationary phase but as Figure 32 demonstrates, other phases may show different selectivity that can help if C18 doesn't do the job. In this example, cardiac drugs were separated on short Rapid Resolution HT columns containing different sub-2 µm packing materials using an isocratic buffered mobile phase consisting of 70% phosphate buffer adjusted to pH 3.0 and 30% acetonitrile.

    First choice of a packing pore size is based on the size of molecules to be analyzed. Typical small molecules can diffuse easily in and out of standard 80 - 120Å pore size packings, but larger peptides and proteins may not. For this reason, it is recommended to use 300Å pore size packings (300SB) for isocratic or gradient separations of peptides and proteins.

    80 - 120Å Packing pore size 300Å

    Small moleculeMW < 2000

    Large moleculeMW > 2000

    Eclipse Plus C184.6 x 150 mm, 3.5 µm

    PN 959963-902

    Poroshell 120 EC-C184.6 x 7.5 mm, 2.7 µm

    PN 697975-902

    Standard analysis

    Poroshell 120 EC-C184.6 x 100 mm, 2.7 µm (superficially porous)

    PN 695975-902

    ZORBAX Rapid Resolution HD Eclipse

    Plus C18, 1200 bar2.1 x 50 mm, 1.8 µm

    PN 959757-902

    Fast analysis

    ZORBAX 300SB-C184.6 x 150 mm, 5 µm

    PN 889395-902

    Standard analysis

    ZORBAX 300SB-C184.6 x 50 mm, 3.55 µm

    PN 8656973-902

    Poroshell 300SB-C182.1 x 75 mm, 5 µm PN 660750-902

    Fast analysis

    Small molecules Large molecules

    A C18 is recommended as the starting column bonded phase for most samples since it maximizes retention for moderately polar to non-polar compounds. Shorter chain phases should be considered if resolution cannot be optimized with a C18 phase or if you are analyzing larger proteins, or very hydrophobic compounds that are difficult to elute from C18 with conventional reversed-phase solvents.

    Eclipse Plus C18 Starting column bonded phase StableBond 300SB-C18

    Figure 31. Reversed-phasechromatography:Overviewforselectingstationaryphase

  • 52

    Essentially, you should choose a phase that matches the requirements of the sample. When working with hydrophobic small molecules, a longer chain alkyl phase such as C18 should be the first choice. If there is too strong of a retention on the C18 phase, then choose a shorter chain C8 or C3 phase. The C8 phases normally have similar selectivity to a C18 phase but show slightly lower retention. For very hydrophobic molecules, it is preferable to use a very short chain phase like C3. If the analyte molecules have aromatic character and can't be sufficiently separated on an alkyl phase, an aromatic bonded phase such as phenyl or diphenyl could be used. In the case where the desired analytes are strongly polar and unretained or slightly retained on a typical reversed-phase chromatography packing, consider hydrophilic interaction chromatography (HILIC) as an alternative HPLC mode (see Section on HILIC, p. 76). If in the course of method development, y